240 likes | 421 Views
Economics of Business. Class 9 Outline. Harvard Extension School Fall 2011 Instructor: Bob Wayland Teaching Assistant: Natasha Wambebe. Cognitive Limits in the Large and Small.
E N D
Economics of Business Class 9 Outline Harvard Extension School Fall 2011 Instructor: Bob Wayland Teaching Assistant: Natasha Wambebe
Cognitive Limits in the Large and Small • Recall that Kaldor, Robinson, Coase, et al recognized that entrepreneurial capacity was limited and ultimately served to limit the scale and scope of the firm • Cognitive limits, coupled with incomplete information about the future also explain many firm failures • Early management scientists didn’t buy it and many still don’t: “We cannot predict the future. But we can create it.” Jim Collins, Great by Design More often than not, companies miss the future not because it was unknowable but because it was disconcerting. Gary Hamel, What Matters Now: How to Win in a World of Relentless Change, Ferocious Competition, and Unstoppable Innovation While I don’t believe that institutional death is inevitable (in theory, every company could be immortal), I Do believe that there are many organizations that deserve to die, and that policy makers should leave them to their fate…. Gary Hamel, Ibid incomplete w/o oral presentation
Previous Episodes of E1600 Who is the decision-maker in the classical model? What super-powers does she have? What factors did the neoclassical model incorporate into the decision-maker’s mental calculus? What is the name for the possible divergence of interests between the “boss” and the “worker”? Why can the worker sometimes successfully evade scrutiny by the boss? How can interests of principal and agent be better aligned? What should guide the decision-making of principals? What is the effect of uncertainty on principal decision-making? Are there constraints to the decision-making capacity of a principal? What role do such constraints, if any, play in determining the size and scope of firms? incomplete w/o oral presentation
Simon’s 1955 Behavioral Model of Rational Choice Some General Features of Rational Choice • The set of alternatives • The relationships that determine the pay-offs (satisfaction, goal attainment…) as a function of the alternative chosen, • The preference-ordering among pay-offs • Importance of determining what is under control Fn: Simon contrasts his model with the “traditional” economic model of economic man. He should have said “classical” model since the super-powers had long been stripped away incomplete w/o oral presentation
Simon’s 1955 Behavioral Model of Rational Choice Components and definitions • “Primitive terms and definitions” that characterize both global and limited models of rational behavior: • A set, A, of behavior alternatives • The subset, Å, of behavior alternative considered by the organism • The set, S, of possible future state of affairs or outcomes of choice • A pay-off function, V(s), defined over all elements s of S. V(s) represents the value or utility placed by the organism on each element s. Simon assume that a cardinal utility for each V(s) is defined and thus the points can be ordered. • Information as to which outcome, s, within S will occur if a particular a in A or Å is chosen.*. • Information as to the probability that a particular outcome s in S will occur if alternative a is chosen. This associates with each element of Sa a probability Pa(s) which is the probability that s will occur if a is chosen. • Pa(s) has all the characteristics of a proper probability function and of course the sum of the probabilities must be 1. *Given incomplete information there may be more than one possible outcome s for a particular behavior a so Simon allows mapping of a to subset Sa – the subset of states that may occur incomplete w/o oral presentation
Simon’s 1955 Behavioral Model of Rational Choice Relationship to game theory incomplete w/o oral presentation
Simon’s 1955 Behavioral Model of Rational Choice Demands of classical concepts of reality: • The game theoretic maxi min rule of selecting the lowest maximum value or least V(s) • Maximize the expected value, using Pa(s), and • The certainty rule, given a known map of a onto sa select the highest pay-off Max V(Sa). • The demands on the choosing organism of these rational choice rules are quite “severe” • The need to attach definite or a definite range of pay-offs to each outcome, • Requiring the ability to specify the exact nature of outcomes, and • Then to be able to order all outcomes. • Simon finds no evidence that choosing organisms have the capability to do so and instead simplify incomplete w/o oral presentation
Simon’s 1955 Behavioral Model of Rational Choice II The Essential Simplifications • Simple pay-off functions • V(s) doesn’t have to be absolutely precise • May represent by only a few general outcomes such as “win, lose, or draw” or “satisfactory and unsatisfactory” • Price above asking and aspirational level is a “win” or “satisfactory” outcome • If sequential offers, take first that equals or exceeds asking; if n-tuples take the highest satisfactory offer among bids incomplete w/o oral presentation
Simon’s 1955 Behavioral Model of Rational Choice Information gathering • Outcome values V(s) may be known but the mapping of the actions in A to the outcomes in S are not. • Simon conjectures that the organism may start with a coarse notion of the A to S mapping, • Then looks at the set of outcomes S for a subset S’ in which all of the outcomes, s, have the simplified V(s) =1 (i.e. are satisfactory) • Then, focusing on that area, S’, he gathers additional information and having refined that mapping, • He looks for an a that will get him there. • The two-stage (four step) approach suggested by Simon is very common among managers faced with mid-value decisions. • Simon suggests parallels with the distinction in linear programming between: • Feasibility testing that determines whether a program is achievable given a set of resources and • Finding an optimal program which has the maximum value among all feasible programs. incomplete w/o oral presentation
Simon’s 1955 Behavioral Model of Rational Choice Partial Ordering of Pay-Offs The classical theory requires that the organism be able to assign to each pay-off a unique place in the order. In practice, Simon suggests that this strict ordering may not be attainable or worth achieving: Group decisions made by people with different preference functions. Multiple decision criteria that are difficult to express with a common denominator. Where an n-dimensional utility surface is not tractable. Where each a maps on a set of n possible s’s, Sa , Bounded apples and oranges-> incomplete w/o oral presentation
Simon’s 1955 Behavioral Model of Rational Choice III. Existence and Uniqueness of Solutions Obtaining a Unique Solution • Global models of rational choice simultaneously examine all alternatives, our limited organisms often examine alternatives sequentially. • The sequential encounters often act as a learning experience. • In many cases, the first satisfactory alternative encountered is chosen. Not a problem seeking check-mate, may be huge problem in selecting a romantic mate • Dynamic factors: • Organism finds satisfactory alternatives are abundant (this may relate to the association of “bounded rationality” with “learning”) • May raise his acceptance or aspirational requirement and conversely, • If it is hard to find satisfactory alternatives, he may lower the bar (everyone looks good at closing time). • Flexibility of the acceptance level ensures that a “satisfactory” solution will be found in most cases. • Simon doesn’t explicitly relate time to number of sequential trials or discuss “analysis paralysis.” incomplete w/o oral presentation
Simon’s 1955 Behavioral Model of Rational Choice Existence of Solutions: Further Possibilities • Instead of changing the bar(s), the organism may expand the candidate set Å. • If few or no acceptable outcomes are found in new, larger Å the organism may expand the area of analysis, • Conversely if an overwhelming number of acceptable outcomes are present, he may shrink the area of analysis. • Simon’s discussion of persistence and the boundary adjustment actions of the organism anticipate to some degree Stigler’s later discussion of the economics of information. • Simon points out that if: • We value our pay-offs in terms of money or utility and • If we knew how to measure the cost of discovering additional alternatives • Then we could create a complete ordering and estimate the optimal degree of persistence. • But the central point of his paper is that the organism doesn’t know those costs. incomplete w/o oral presentation
Simon’s 1955 Behavioral Model of Rational Choice Simon makes an intriguing point: • “The question of how to behave ‘rationally’ given these limitations, is distinct from the question of how its capabilities could be increased to permit action that would be more ‘rational’ judged from the mountaintop of a more complete model.” • “…judged in terms of the survival value of his choice mechanism.” • Seems to be distinguishing between the organism’s given capabilities and set of choice mechanism and the adaptive efforts of learning that might make the organism more “fit “for a broader range of environmental selection pressures. (relationship to Alchian points, Gibbons…?) incomplete w/o oral presentation
Simon’s 1955 Behavioral Model of Rational Choice IV. Further Comments on Dynamics Pay-offs may depend on previous decisions as well the instant one “It doesn’t know how well it likes cheese until it has eaten cheese.” Trail and error is a learning process in which early consequences condition later choices. incomplete w/o oral presentation
Simon’s 1955 Behavioral Model of Rational Choice V. Conclusion Paradox of an organism of limited knowledge and capability making choices that diverge from reality may explain many of the phenomena of organizational behavior Perfectly rational Vulcans, might all get along, make consistently excellent decisions, and have little need for complex contracting and governance structures Simon drew connection to potential computer designs based on real decision making behavior incomplete w/o oral presentation
Bounded Rationality and Organizational Learning (1991) The Structure of Roles • Organizations are best seen as systems of interrelated roles • Distinguishes between prescribed behavior systems and prescribed decision premises, the latter defining roles. • “Roles tell organization members how to reason about the problems and decisions that face them; • where to look for appropriate and legitimate informational premises and goal (evaluative) premises, and what techniques to use in processing these premises.” • Compare to notion of routines as genes used by Nelson and Winter. • Each individual role presumes that appropriate and communicating roles surround it and together they all form an organization or role system. incomplete w/o oral presentation
Bounded Rationality and Organizational Learning (1991) Organizational Learning and Innovation • An organization that desires to differentiate itself from the packaged disciplines taught at universities has to perform the education or re-education itself. • Turnover is beneficial if it removes barriers to the new posture but costly if it increases the re-education costs. • Some organizations such as the Forest Service see their values and processes taught and inculcated in Schools of Forestry. • Roles, i.e. decision premises are hardy. • A resident organization decision premise or behavior resist change from new entrants if they enter at relatively low rates • Smokey the Bear mentality difficult to change to accept controlled burns • Important to understand “mechanisms that can be used to enable an organization to deviate from the culture in which it is embedded.” • Simon calls for more case studies and, strangely, invokes Darwin’s study of the Galapagos Islands as a “case study.” incomplete w/o oral presentation
Bounded Rationality and Organizational Learning (1991) Organizational Memory • Most organizational memory is in heads, some in computer banks. • Turnover is the enemy of long-term organizational memory. Erosion of institutional memory can be a good or a bad thing. • Cognitive psychology has learned a lot about expertise: • It is based on extensive knowledge (parents vs. teenagers) • World class expert has approximately 50,000 chunks of relevant information stored in his head’s indexed encyclopedia • Literally no one becomes an expert in a professional domain without at least ten years of full time learning. • Based on these notions, Simon suggests that an organization can also be represented as a collection of production systems and suggests this is more true as automated expert systems evolve. • His discussion begs but doesn’t address how the cues work in an organization. incomplete w/o oral presentation
Bounded Rationality and Organizational Learning (1991) Ingesting Innovations From Without Recall Alchian’s point about imitation being an important source of innovation and the analogue to biological heredity. Recall also Klepper’s surveillance function, F, that enabled firms to quickly adopt and incorporate innovations from other firms. Simon appreciates the importance of assimilating ideas from outside the organization and transmitting it to the point of implementation. Simonadmitsthat university research does not immediately benefit students and most of the knowledge imparted to students arose elsewhere. Way to ensure that professors stay mentally alive. Think of much of the research activity as “intelligence gathering” akin to that term’s military use. Too much concentration on producing only new knowledge can foster NIH and re-inventing the wheel. incomplete w/o oral presentation
Bounded Rationality and Organizational Learning (1991) R&D and Manufacturing Much noted problem of developing new products that can be produced and taken to market within the constraints of the organization The various organizational units have their distinct production systems and expertise. Finding a path for the new product from idea to market involves satisfying each relevant system. Recall the bounded areas for apples and oranges in first reading. Consider creative destruction and the observed failure of many to recognize threats (disruptive technologies) from outside incomplete w/o oral presentation
Bounded Rationality and Organizational Learning (1991) Acquiring New Problem Representations • Simon cites research on “how people solve problems by searching selectively through a problem space defined by a particular problem representation.” (could he be referring to his own?) • He distinguishes two cases of acquiring new representations; • The learner, often a new member, is presented with the correct representation but has to learn how to use it, and • The organization faces a new problem and creates a new problem representation and equips members to use it incomplete w/o oral presentation
Bounded Rationality and Organizational Learning (1991) Creating an Organization • People bring their old organization ideas with them to new situations. Why Representation Matters (how you see a problem) • Bounded rationality seems closely related to problem representation. • Policy statements imply representations. • USS shifted business portfolio from almost all steel to both oil and steel and became USX • USX moved its problem representation accordingly incomplete w/o oral presentation
Next Week Readings: Williamson (1979) on relationships Optional: May cover supplier relationships and equity, time permitting Don’t forget to check Piazza for midterm info and to use it for intra-class discussion incomplete w/o oral presentation